Biological Variation for the Interpretation of Cardiac Biomarkers in the Acute Setting

A 70-year-old white male presents to the emergency department (ED) with dyspnea and a 25-minute history of chest pain. He has a medical history of congestive heart failure and renal insufficiency. His electrocardiogram was unremarkable for new onset myocardial ischemia. Several weeks ago, the patient was seen at his primary care physician's office for a routine checkup. Blood was collected and tested for cardiac troponin using a novel high sensitivity assay (hs-cTn) which produced a result of 4.5 pg/mL (99th percentile limit less than10 pg/mL). Using the same manufacturer of the hs-cTnI assay, the patient's result on blood collected at ED presentation was 9.0 pg/mL. The B-type natriuretic peptide (BNP) was 433 pg/mL (cutoff 100 pg/mL) and the serum creatinine was 1.5 mg/dL (normal 0.7-1.4 mg/dL). A subsequent sample collected three hours after ED presentation produced a hs-cTnI of 15 pg/mL (Fig. 1).

Co-morbid cases such as this present diagnostic challenges to ED physicians and cardiologists because the initial troponin result was normal and the subsequent value was only marginally above the 99th percentile. Increased concentrations of BNP and creatinine were consistent with the history of heart and renal disease in this case. The presence of these diseases alone can be responsible for the abnormal troponin level observed. Of immediate concern, however, was the increasing level of cardiac troponin upon repeat draw. While there is evidence that the patient has suffered some acute myocardial injury, it is not clear if the release is due to exacerbation of pre-existing chronic diseases. How can the change in troponin results be used to determine the etiology of release?

In order to interpret results of cardiac biomarker results in disease, it is important to understand how changes in clinical laboratory test results occur during health. The statistical science of "Biological variation" refers to variances between individuals (coefficient of variance, CVG) and within an individual (CVI) over time. There is also analytical variances (CVA) that exist within the measurement process.1 It is important that clinical laboratories keep the CVA to a small fraction of the intra- and inter-individual variances.2

Biological variation studies are useful to determine the appropriateness of the upper reference limit. International guidelines have determined that the cutoff concentration for cTn assays is established at the 99th percentile of a healthy population. The ratio of intra-individual to inter-individual variation (CVI/CVG), termed the "index of individuality" is used to determine the appropriateness of this strategy. Reference ranges are useful for tests by which there are little variances between individuals relative to the population as a whole. The serum sodium concentration is consistently between 135 and 145 mmol/l, irrespective to the age, gender, body mass index and ethnicity of tested subject resulting in a high index of individuality. Reference ranges are appropriate for sodium and other tests high indices of individuality.2

For cardiac troponin, there is significant difference between individuals relative to within an individual producing a low index of individuality. This means that a singleFigure 1: Biological Variation for the Interpretation of Cardiac Biomarkers in the Acute Setting value cannot be effectively compared to a population-based normal range. The optimum interpretation of these test results would be to use the individual's own normal range, determined by repeated analysis. Figure 1 illustrates this point. Repeated hourly cTn results from a dozen healthy individuals are displayed. The data shows that within an individual, there is only small changes in cTn concentrations (CVI= 10%). The heart does not remodel over a short duration in health and does not release a substantial amount of troponin. However, there is a greater than 5-fold difference in cTn levels between the individuals tested (CVG = 57%). Some of this may be related to differences in troponin concentrations between gender and different subject age.3 By definition, the 99th percentile population-based cutoff limit must incorporate all of these results (i.e., a cutoff of 10 pg/mL in this case). This designation can mask a clinically significant change of troponin values that may still be below the diagnostic cutoff concentration, as observed in this case. It would be ideal to apply an individual's own "normal range" for tests such as troponin that have a low index of individuality,4 if such baseline data were available.

The intra-individual variance of a laboratory test is also used to calculate the "reference change value" (RCV), an important attribute for tests that have a low index of individuality whereby serial testing becomes essential. The RCV is calculated from an appropriate healthy reference population, and determines the maximum change in biomarker test results that is expected during health (details of the specific mathematical equations are presented elsewhere).2 When serial biomarker tests are used in clinical practice, such as for cTn, changes in results due to disease must exceed the test's inherent RCV. In the case of troponin, the RCVs have been calculated to be 50% (when calculated relative to the higher value).5 For this case report, the second troponin result in the ED was 67% greater than the admission value (15 vs. 9.0 pg/mL), just above the RCV for this assay. However, the admission value was 100% greater than the patient's baseline value (4.5 pg/mL), despite the fact that both were within the 99th percentile. These serial cTn values exceed the RCV indicating that these changes exceed limits observed during health (Fig. 2).

Knowledge of both the baseline level and the second value at three hours after ED presentation provides evidence of acute myocardial injury. This patient wouldFigure 2: Biological Variation for the Interpretation of Cardiac Biomarkers in the Acute Setting warrant further medical evaluation for the presence of acute coronary disease. Values exceeding cTn RCV threshold do not necessarily establish a diagnosis of acute coronary syndromes. The establishment of the optimum cutoff for serial change at or above the RCV will require evaluations through clinical trials. With the release of high-sensitive troponin assays, establishing serial cutoff limits should be the focus of future biomarker studies in ACS.

The determination of the biological variation for cTn assays has only been conducted within the last few years with the development of high-sensitivity assays.6 The prior generation assays were insufficiently sensitive to reliably measure troponin in blood of healthy subjects. With the use of high-sensitive assays, it is possible to establish a "baseline" level for patients at high risk for cardiovascular disease. Using the biomarker's "homeostatic setpoint" another attribute of a biomarker derived from biological variation studies,2 approximately six repeated samples are required to establish a mean baseline troponin value. The concept of interpreting biomarker results in an acute setting to a subject's own baseline value requires that testing is conducted using the same assays (from primary care and testing conducting in the hospital where the patient is taken) and biomarker (troponin T vs. I). For the moment, this is not possible today because there is no standardization between commercial cardiac troponin I assays, so that results between assays do not agree. This is an issue that is being addressed by a Committee of the International Federation of Clinical Chemistry, despite the recognized difficulties of achieving this goal.7 This objective can be achieved for cardiac troponin T, as there is standardization for this assay and the biological variation has been established.8 However, troponin T testing is limited to one manufacturer, and thus this biomarker is not as widely used in clinical practice as cardiac troponin I.


  1. Wu AHB. Biological and analytical variation of clinical biomarker testing: implications for biomarker-guided therapy. Curr Heart Fail Rep 2013;10:434-40.
  2. Fraser CG. Biological variation: from principles to practice. Am Assoc Clin Chem Press, Washington, DC: 2001.
  3. Venge P, Johnston N, Lindahl B, James S. Normal plasma levels of cardiac troponin I measured by the high-sensitivity cardiac troponin I Access prototyop assay and the impact on the diagnosis of myocardial infarction. J Am Coll Cardiol 2009;54:1165-72.
  4. Wu AHB, Christenson RH. Analytical and assay issues for use of cardiac troponin testing for risk stratification in primary care. Clin Biochem 2013;46:969-78.
  5. Wu AHB, Lu A, Todd J, Moecks J, Wians F. Short- and long-term biological variation in cardiac troponin I with a high-sensitivity assay: implications for clinical practice. Clin Chem 2009;55:52-8.
  6. Nordenskjold AM, Ahlstrom H, Eggers KM, Frobert O, Jaffe AS, Venge P, Lindahl B. Short- and long-term individual variation in cardiac troponin in patients with stable coronary artery disease. Clin Chem 2013;59:401-9.
  7. Christenson RH, Bunk DM, Schimmel H, Tate JR. on behalf of the IFCC Working Group on Standardizaiton of Troponin I. Put simply, standardization of cardiac troponin I is complicated. Clin Chem 2012;58:165-8.
  8. Frankenstein L, Wu AHB, Hallermayer K, Wians FH, Giannitsis E, Katus HA. Biological variation and reference change value of high-sensitivity troponin T in healthy individuals during short and intermediate follow-up periods. Clin Chem 2012;57:1068-71.

Keywords: Chest Pain, Creatinine, Dyspnea, Electrocardiography, Heart Failure, Myocardial Ischemia, Natriuretic Peptide, Brain, Physicians, Primary Care, Renal Insufficiency, Troponin I

< Back to Listings